Influence of the Dopant Concentration on the Photoelectrochemical

Mar 22, 2018 - The experimental dependencies of photoelectrochemical efficiency on Al dopant concentration indicate that there is an optimal Al concen...
0 downloads 0 Views 680KB Size
Subscriber access provided by Queen Mary, University of London

C: Physical Processes in Nanomaterials and Nanostructures

Influence of the Dopant Concentration on Photoelectrochemical Behavior of Al-Doped TiO

2

Anna A. Murashkina, Aida V. Rudakova, Vladimir K Ryabchuk, Konstantin V. Nikitin, Ruslan V. Mikhaylov, Alexei V Emeline, and Detlef W. Bahnemann J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b12840 • Publication Date (Web): 22 Mar 2018 Downloaded from http://pubs.acs.org on March 22, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Influence of the Dopant Concentration on Photoelectrochemical Behavior of Al-doped TiO2 Anna A. Murashkina*, Aida V. Rudakova, Vladimir K. Ryabchuk, Konstantin V. Nikitin, Ruslan V. Mikhailov, Alexei V. Emeline*, Detlef W. Bahnemann Laboratory “Photoactive Nanocomposite Materials”, Saint-Petersburg State University, Ulyanovskaya str. 1, Peterhof, Saint-Petersburg, 198504 Russia

*Corresponding Authors E-mail: [email protected] (A.A.Murashkina), [email protected] (A.V.Emeline)

Abstract In the present study we explored the effect of Al dopant concentration within the range < 1.1 wt% on the photoelectrochemical activity of an Al-doped TiO2 photoanode. The experimental dependencies of photoelectrochemical efficiency on Al dopant concentration indicate that there is an optimal Al concentration of 0.5 wt.% corresponding to the highest photoactivity. The analysis of the spectral dependencies of the photocurrent confirms that 0.5 wt.% of Al provides the highest activity at photoexcitation in both intrinsic and extrinsic absorption spectral range. It was also shown that Al doping does not affect the optical band gap of TiO2. The dependence of photoelectrochemical activity on Al concentration correlates with corresponding dependencies of the flat band potential and work function indicating the Fermi level shift toward the conduction band for the Al concentration < 0.5 wt.% and toward the valence band for the Al concentration > 0.5 wt%. Such alteration of the Fermi level position is explained in terms of alteration of the type oo / of major compensating intrinsic defects from [ VO - Ti Ti ] for Al concentration < 0.5 wt.% acting / oo / × as shallow traps, to [Al Ti − VO − Al Ti ] for Al concentration > 0.5 wt.% acting as deep traps.

Transformation of compensating defects from shallow traps, which are ineffective in charge recombination processes to deep traps which act as effective recombination centers is responsible for the optimal dopant concentration, 0.5 wt.%, to achieve the higher photoelectrochemical activity of Al-doped TiO2.

Introduction

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Titanium dioxide is the most extensively studied photoactive material since the discovery of the of water photo-assisted electrolysis with titanium dioxide photo-anode.1 Materials based on titanium dioxide can be used in photocatalytic systems for water and air purification, formation of self-cleaning and antibacterial coatings, in solar cells and sensors.2-5 Titanium dioxide possesses a sufficiently high photochemical (photocatalytic) activity, good chemical stability and non-toxicity and low cost of its production. However, the band gap energy of TiO2 (about 3.0 and 3.2 eV for rutile and anatase, respectively) is rather large. Therefore, the sensitization of the titanium dioxide toward visible light is one of the major tasks for increasing the efficiency of photocatalytic and photoelectrochemical systems. An effective way of the spectral sensitization is doping and co-doping of titanium dioxide with metal and nonmetal.6-9 Most researchers attributed the sensitization of titanium dioxide caused by doping to the formation of "impurity sub-bands" in the energy spectrum of TiO2 (so-called "band gap narrowing") and this fact is often confirmed by the results of the theoretical calculations, in particular, for nitrogen doped titanium dioxide.10 According to an alternative point of view, the absorption of doped titanium dioxide is due to intrinsic defects arising as a result of the charge compensation of the heterovalent impurities.11 At the same time, aliovalent doping of TiO2 could lead to higher recombination losses of photocarriers.12 The effect of aluminum doping on the photocatalytic activity was investigated in several studies.13-15 It was found that Al3+-ion is introduced into the titanium oxide through substitution. However, the incorporation of aluminum in the rutile lattice by interstitial dissolution is also possible.16 The properties of TiO2 thin film electrodes depend on the conditions of polycrystalline structure formation (the original components, temperature, etc.),17 controlled and uncontrolled impurities. New energy levels appearing in the process of doping can act as traps for charge carriers affecting the charge transport processes in semiconductors18 and therefore, the photocurrent in PEC systems.19 Introduction of acceptor impurities causes the formation of oxygen vacancies as compensation defects, which can act as electron traps. Otherwise, the formation of hole centers induced by UV-irradiation was observed by EPR technique.20 Applying the density functional theory (DFT) calculations it was shown that holes are strongly localized in the O2p nonbonding orbital of a three-coordinated O ion. Accordingly, the trapping of photogenerated electrons and holes by intrinsic defects affected by dopants reduces the mobility of charge carriers. Aside from that, the formation of electronic states of new defects and

ACS Paragon Plus Environment

Page 2 of 19

Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

impurities within the band gap of semiconductor can significantly affect the position of the Fermi level or the flat band potential of the electrode material. The shift of the flat-band potential to the more negative potentials’ region promotes, for instance, the process of water photolysis should the Fermi level reach the level of the electrochemical potential of hydrogen evolution. The aim of this study is to explore the photoelectrochemical activity of aluminum-doped titanium dioxide as a function of the Al-dopant concentration to establish the role of the dopant in alteration of the electronic behavior of titania. Al-dopant concentration was varied within the range 0.0 -1.1 wt.%.

Experimental 1. Electrode preparation Polycrystalline x-Al-TiO2 (x, Al wt% ‒ 0.0, 0.1, 0.3, 0.5, 0.7, 0.9, 1.1) electrodes were made by sol-gel dip-coating deposition (KSV Nima dip coater) on ITO glasses (Aldrich, 15-25 Ω/cm). x-Al-TiO2 sols were prepared using titanium isopropoxide (TTIP, 97%, Sigma-Aldrich), aluminum isopropoxide (ATIP, 99.98%, Acros Organics), isopropanol (≥ 99.0%, Vekton) and citric acid dehydrate (CADH, 98.0%, Vekton) as titanium and aluminum precursors, solvent and complexing agent, respectively. Molar stoichiometry of metal ions to citric acid dehydrate was taken as 1:3. To form a good contact with conductive glass the first four layers were formed as a dense thin film by using the slow withdrawing velocity 10 mm.min-1 and the drying procedure for 30 min at 200 °C after each dipping. To make the rough surface of electrodes with higher surface area, the next six layers were formed with higher withdraw velocity 100 mm.min-1 followed by annealing at 500 °C in the air for 5 hours. 2. Electrode characterization X-ray diffraction measurements with Bruker «D8 DISCOVER» high-resolution diffractometer (CuKa X-ray radiation, within the angle range of 20° ≤ 2θ ≤ 80° with a scanning speed 5.0°/min) were used for the crystal phase determination of all studied x-Al-TiO2 samples. Structural reference data were taken from ICSD database. According to XRD data, pure and Aldoped TiO2 electrodes were crystallized in anatase phase (see Figure S1 in the Supporting Information section). No formation of any aluminum-containing phase was observed for all synthesized samples.

ACS Paragon Plus Environment

The Journal of Physical Chemistry

XPS analysis (Escalab 250Xi spectrometer, Thermo Fisher Scientific Inc., USA) confirms the incorporation of Al3+ cations into the TiO2 lattice (see Figure 1a). The survey XPS scan as well as narrow scans for Ti4+, Al3+ and O2- states are presented in the Figure S2 in the Supporting Information section. It is remarkable that no new titanium states were detected while less negative oxygen states and their dependence on the aluminum doping (see below in Results section) were observed (see Figure 1 b, c). 0,04

Al 2p

1.1

0.5

∆ (IAl-Ipure)

Intensity, a.u.

Intensity, a.u.

O 1s

0,02

0,00 534

532 Binding Energy, eV

530

1.1 0.5

0.1 65

70

75

80

85

0.1 0 534

532

530

528

526

Binding Energy, eV

Binding Energy, eV

b

a

10000

SOirr, a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 19

8000

6000

4000 0,0

0,2

0,4

0,6

0,8

1,0

1,2

x, wt. % Al

c

Figure 1. a) XPS spectra of Al3+ 2p states in Al doped TiO2 b) XPS spectra of oxygen O1s states in Al doped TiO2 and their deconvolution in two different forms (blue band relates to the regular oxygen states and red band corresponds to less negative (irregular) oxygen states) c) dependence of the irregular oxygen concentration on the Al dopant concentration.

The surface morphology of all titania electrodes was explored by scanning electron microscopy (Zeiss Supra 40 VP system) (see Figures S3 in the Supporting Information

ACS Paragon Plus Environment

Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

section). The electrode images confirm that ITO glasses are completely covered by x-Al-TiO2 layers. The particle size of titania polycrystalline coatings was about 10 nm. It was shown that the flakes of agglomerated particles were formed for all samples that made the electrode surfaces sufficiently rough. Work function measurements were performed with scanning Kelvin probe system SKP5050 (KP Technology) versus a golden reference probe electrode (probe area 2 mm2). The probe oscillation frequency was 74 Hz, the backing potential was 7000 mV. Work function values were obtained by averaging of 50 data points for five different sites of each sample.

3. Photoelectrochemical experiments Photoelectrochemical measurements were performed at room temperature using threeelectrode cell with a quartz window, Ag/AgCl reference electrode (0.222 V vs NHE potential) and Pt wire as a counter electrode. K2SO4 aqueous solution (0.2 M) was used as an electrolyte. Electrochemical measurements were run with computer-controlled Reference 600 (Gamry Instruments, Inc) potentiostat. Current-voltage dependences (I-V curves) were recorded in the voltage region from -1 V to +1 V using a scan rate of 30 mV/s. 150 W Xe lamp (LOMO) was used as a light source. The dependences of photocurrent on intensity and wavelength of acting light were carried out with the setup consisting of a 1000 W Hg-Xe lamp (Oriel) with a water filter and MDR-12 monochromator (LOMO) and a set of color filters (Vavilov State Optical Institute). The spectral resolution of the wavelength dependence measurements was about ∆λ = ±2.5 nm. The intensity of monochromatic light was measured with NOVA II light power meter (Ophir). The dependences of photocurrent on light intensity were measured within the range of 0−4.0 mW.cm-2 (see Figure S4 in the Supporting Information section). It was shown that the dependences are linear, which indicates that the efficiency of the photoelectrochemical processes is limited only by the rate of photocarrier generation.

Results Figure 2 demonstrates the current-potential curves for pristine and 0.5 wt.% Al-doped TiO2 thin film electrodes in the dark and under broad band (Xe-lamp) irradiation Note, that the photocurrent attains the saturation at a significantly positive potential that indicates that at given

ACS Paragon Plus Environment

The Journal of Physical Chemistry

experimental conditions the photoelectrochemical process is limited by photogeneration of charge carriers in TiO2.21 The current-potential curves for all samples are presented in Figure S5 (Supporting Information section) and demonstrate a significant increase of current density with increase of dopant concentration up to x = 0.5 wt.% followed by a decay of photocurrent at higher dopant concentration.

0,1

j, mA/cm2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 19

0,0

-0,1

-1,0

-0,5

0,0

0,5

1,0

E, V vs Ag/AgCl

Figure 2. Сurrent-voltage dependences for pristine (square) and 0.5wt.% Al-doped TiO2 (circle) photoanodes in dark (open symbols) and under illumination (filled symbols).

The efficiency of the photoelectrochemical conversion (water photoelectrolysis) was calculated by equation 1 (see ref.22):

η=

j ph × (1.229 − Vapp )

P×S

× 100% .

(1)

where IPh is a photocurrent, Vapp is an applied potential (V), P is a radiation power density (Wm−2), S is an irradiated electrode area (m2) and 1.229 is the thermodynamic waterdecomposition potential (V). The dependences of the estimated efficiencies of the water photoelectrochemical conversion on the applied potential for all studied samples are shown in Figure 3.

ACS Paragon Plus Environment

Page 7 of 19

0,5 0.0 0.1 0.3 0.5 0.7 0.9 1.1

0,4

η c, %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

0,3 0,2 0,1 0,0 -0,9

-0,6

-0,3

0,0

0,3

0,6

E, V vs Ag/AgСl

Figure 3. Photoconversion efficiency as a function of the applied potential for TiO2 thin film photoanodes.

As evident from the presented dependences (see Figure 3), the efficiency of water photodecomposition increases from the open circuit potential, goes through the maximum and then decreases at the thermodynamic potential of the water decomposition yielding the formation of molecular hydrogen for all samples. Remarkably, the conversion efficiency (ηc) of doped samples increases to x = 0.5 wt.% and then decreases at higher dopant concentrations within the whole range of applied potentials. Moreover, one can distinguish two noticeable extrema around 0 V and -0.6 V with significantly different behavior. The extremum around 0 V is well pronounced and increases with increase of the Al concentration up to x = 0.5 wt.%, and at higher dopant concentrations it completely disappears. At the same time, the extremum around -0.55 V is not observed at low doping value x = 0.1 wt.% and appears and remains observable only at higher Al concentration (x ≥ 0.3 wt.%). It indicates a significant redistribution of the electronic states in titania caused by Al doping. Figure 4 shows the dependence of the efficiency of photoelectrochemical water conversion, ηc, on the Al dopant concentration at the applied potential – 0 V vs Ag/AgCl. The obtained dependence goes clearly through the maximum at x=0.5 wt.%.

ACS Paragon Plus Environment

The Journal of Physical Chemistry

0,6

0 V vs Ag/AgCl

0,4

η c, %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 19

0,2

-0,55 V vs Ag/AgCl

0,0 0,0

0,2

0,4

0,6

0,8

1,0

1,2

x, wt.%Al

Figure 4. The dependence of the conversion efficiency (ηc) on Al dopant concentration.

Thus, the Al doping significantly changes the efficiency of titania in water photoelectrolysis process and provides the maximum efficiency at Al dopant concentration x = 0.5 wt.%. Flat band potentials Vfb of the undoped and doped samples were determined from the photocurrent measurements according the procedure described elsewhere.23 The applied external potential Vapp affects both the dark-current and the photo-current. Upon the illumination of semiconductors the interfacial band bending decreases. The photo-current approaches the darkcurrent near the flat band potential Vfb and graphically, the crossing point of j-V curves corresponds to Vfb. In addition to the photoelectrochemical method we also apply the Mott-Shottky method to determine the flat band potential values for the pristine and Al-doped TiO2 electrodes (see Figure S6 in the Supporting Information section). The obtained data and the dependence of the flat band potential on the Al-dopant concentration corresponds well to those measured using the photoelectrochemical approach (see data in the Table 1 and Figure S7 in the Supporting Information section). The values of the flat band potential Vfb (V vs Ag/AgCl) are shown as a function of Aldopant concentration in the Figure 5 and presented below in Table 1. The dependence of the flat

ACS Paragon Plus Environment

Page 9 of 19

band potential on the Al-dopant concentration correlates well with the alteration of the work function (measured by Kelvin probe method) of the corresponding electrodes (see Figure 5 and Table 1 below). In other words, smaller flat band potential correlates with the smaller value of the work function. In turn, the more negative potential (and the smaller value of the work function) corresponds to the higher position of the Fermi level with respect to the vacuum level.

5,1

Work function, eV

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

5,0

4,9

4,8

0,0

0,2

0,4

0,6

0,8

1,0

1,2

x, wt.% Al

a

b

Figure 5. The dependence of the flat band potential (a) and the work function (b) of all studied electrodes on the Al dopant concentration. Remarkably, both dependences of the flat band potential and work function correlate well with the dependence of the efficiency of water photoelectrolysis shown in the Figure 3 and prove the assumption made earlier about the redistribution of the electronic states in titania caused by Al doping. Figure 6 shows the spectral dependences of incident photon-to-current conversion efficiency (IPCE), η(λ), for nominal pure and Al-doped TiO2. The IPCE values were calculated using the following equation (2)24 IPCE(%) =

1239.8 ⋅ jph I mchr ⋅ λ

× 100% ,

(2)

where 1239.8 (in V.nm) is the product of Plank’s constant, and the speed of light; jph (mA.cm-2) is the stationary photocurrent density that was taken from chronoamperometry measurements;

ACS Paragon Plus Environment

The Journal of Physical Chemistry

Imchr (mW.cm-2) is the power intensity of acting monochromatic light; and λ (nm) is the wavelength of the monochromatic light. Insertions in Figure 6 demonstrate the dependences of IPCE on the Al-dopant concentration for both intrinsic (λ = 365 nm) and extrinsic (λ = 420 nm) spectral regions of photoexcitation. Note, that both dependences demonstrate the maxima at Al-dopant concentration 0.5 wt.%, while at higher dopant concentration, the IPCE significantly decreases. Thus again, there exists an optimal concentration of Al-dopant around 0.5 wt.% for higher IPCE of TiO2 in both intrinsic and extrinsic absorption spectral regions of titania. 5 0,15

4

4

4

η, %

3

η, %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 19

2

2

3

0,10

420 nm

365 nm 2

1

1 6

3

0,05 1

7 0 350

5 400

450

500

0 0,0

wavelength, nm

0,2

0,4

0,6

0,8

1,0

0,00 1,2

x, wt% Al

a

b

Figure 6. Dependences of photocurrent quantum efficiency on wavelengths of the actinic light for Al(x) doped TiO2 electrodes (1 – x = 0.0 wt.%, 2 – x = 0.1 wt.%, 3 – x = 0.3 wt.%, 4 – x = 0.5 wt.%, 5 – x = 0.7 wt.%, 6 – x = 0.9 wt.%, 7 – x = 1.1 wt.%, (b) the dependence of IPCE on Al-dopant concentration at 365 and 420 nm.

The band gap energies Eg of Al-doped titanium dioxide were determined from the photocurrent spectra (see Figure S8 in the Supporting Information section) by making an assumption that the photocurrent at the edge of the fundamental absorption was proportional to the absorption coefficient. Then, the band gap energies, Eg, can be found from the spectral dependences of IPCE by their re-plotting in the form (Iphhν)m as a function of the photon energy (hν) in the vicinity of the band gap:

ACS Paragon Plus Environment

Page 11 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(Iphhν)m ≈(hν - Eg)

(2)

The band gap energies (Eg, eV) estimated for indirect band-to-band optical transition (m = 0,5) for all tested samples are in the range 3.11-3.13 eV (Table 1), that infers that Al-doping does not change the band gap energy of TiO2. Table 1. x, wt %

0.0

0.1

0.3

0.5

0.7

0.9

1.1

Eg, eV

3.13±0.01

3.11±0.01

3.11±0.01

3.12±0.01

3.13±0.01

3.11±0.01

3.13±0.01

-0.64±0.01

-0.6±0.02 -0.74±0.02 -0.81±0.02 -0.75±0.01 -0.71±0.01 -0.60±0.01

Vfb, V (J-V) Vfb, V (MottShottky)

-0,71±0.01 -0,66±0.01 -0,71±0.02 -0,76±0.01

-

-0,72±0.01 -0,71±0.02

Discussion According to the obtained data the photoelectrochemical performance depends on the Al dopant concentration. At low level of doping (≥ 0.5 wt.%) the improvement of the photoresponse occurs both in the intrinsic and extrinsic TiO2 absorption regions. With increase of the dopant concentration, the photoefficiency of Al-TiO2 electrodes goes down (see Figures 1 and 2). At the same time, the alteration of photoelectrochemical activity as a function of Al dopant concentration correlates with the behavior of the flat band potential, work function and density of the lower negative oxygen states. In other words, all these dependences correlate with each other. That suggests that Al doping of TiO2 is a major cause of the alteration of electronic state distribution in TiO2, which results in the observed consequences: alteration of oxygen states, work function and therefore, the photoelectrochemical activity. Remarkably, all experimental dependencies are not linear and demonstrate extrema at Al concentration 0.5 wt.%. That means that Al doping affecting the electronic state distribution changes not only a quantity of the electronic states but also the type of the electronic states. Indeed, the observed positive shift of the flat band potential and negative shift of the work function for the samples with Al dopant

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 19

concentration 0.1 – 0.5 wt.% and 1.1wt%) suggests the displacement of the Fermi level towards the conduction band (CB), while for other samples with higher Al dopant concentration the Fermi level is shifted closer to the valence band (VB). This trend allows us to suggest the formation of one type of defect states near the conduction band for 0.1 – 0.5 wt.% Al-TiO2 and of another type of defect states closer to the valence band

for 0.5 – 1.1 wt.% Al-TiO2,

respectively. Particularly, this is illustrated by the alteration of XPS signal of less negative oxygen states (see Figure S2 in the Supporting Information section). /

According to the defect chemistry, the substitution of Ti4+ with Al3+ ( Al Ti ) results in the formation of point defects due to the charge compensation effect since the deficiency of the positive cation charge brought by Al3+ doping must be compensated by intrinsic positively charged defects. Kuznetsov et al.11 identified a number of absorption bands with peaks at about 2.9 eV, 2.5 eV and 2.0 eV attributed to intrinsic defects, whose appearance was induced by various impurities including Al. Such defects in TiO2 were described as F-type color centers, i.e. anion vacancies with one (F+-centers) or two (F0-centers) localized electrons. The wide peak in the spectral dependence η(λ) (see Figure 1b) at about 420 nm is similar to one of the absorption bands attributed to F-centers [11]. Indeed, the substitution of Ti4+ ions by Al3+ ions into the oo lattice of titanium oxide can result in the formation of anion vacancies ( VO ) (according to

equation 2): 2 ( Al )

/ TiO 2 → 2Al Ti + VOoo + 2O ×O .

(2)

Thus, the electrons localized at the oxygen vacancies form stable localized states giving a photoresponse within the visible spectral range.25 At the same time, Al ions occupying the substitutional positions in the TiO2 lattice and creating a charge imbalance in the crystal lattice can be compensated by hole states which are strongly localized in the O2p nonbonding orbital of a three-coordinated O ion.20 Thus, the Al3+ dopant ion in the Ti-position can stabilize the hole trapping centers localized at 2.23 eV above the top of the VB.20 However, as well known, in the TiO2 crystal lattice the electron density can be localized on Ti4+-ions which are neighboring to oxygen vacancies.26 The oxygen vacancies and the oo / associated Ti3+ ions are formed as VO ~ Ti Ti dipole pairs, and the dipole pairs were aligned

ACS Paragon Plus Environment

Page 13 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

opposite to the direction of the intrinsic dipole moments, and the amount of the dipole pairs, which will counteract the intrinsic dielectric dipoles, increases with increasing of the dopant concentration.26 One oxygen vacancy induces the formation of three low-coordinated (five-fold) Ti ions. Consequently, the two electrons may be both delocalized on several Ti ions or localized on single Ti ions. Accordingly, the formation of Ti3+ might be explained in terms of electron localization of on the regular cationic sites - Ti4+. The energy levels of such defects lie at the bottom of the CB.27 This supposition is confirmed by the presence of the Ti3+-peak about -0.55 V vs Ag/AgCl in the j-V-dependence (Figure 2) as well as the wide peak at about 420 nm on η(λ) in Figure 1 that belongs to the absorption of the these defects. Hence, in general the Al doping can result in a wide range of different types of intrinsic defects corresponding to different electronic states within the band gap of TiO2. Nevertheless, the observed dependencies of the flat band potential and work function on the Al dopant concentration imply that there is a certain distribution of the electronic states for each dopant concentration and redistribution of the electronic states from one sample to another is dictated by the dopant concentration. In a simplified scenario the number of compensating intrinsic defects, such as oxygen vacancies, should scale linearly with Al-dopant concentration. However, one should take into consideration that if each dopant cation Al3+ bringing the excess of negative charge yields the formation of a single oxygen vacancy possessing double positive charge the complete charge balance can be achieved only provided that another intrinsic defect such as a shallow electron trap Ti3+, is also formed. Thus, incorporation of Al3+ into the lattice results in stabilization of oxygen vacancies and shallow electron traps. In turn, this leads to a decrease of the work function and increase of the flat band potential that is a shift of the Fermi level toward the bottom of conduction band (see Figure 7). Clearly, this supposition can be an explanation of the observed dependencies for both parameters on the Al dopant concentration in the range 0 – 0.5 oo / wt%. In other words, at low Al dopant concentration the compensating defects are [ VO - Ti Ti ]

acting as a shallow traps. Note that in terms of kinetic behavior of photogenerated charge carriers, the shallow traps are ineffective recombination centers. Also the formation of shallow electronic defects helps electron migration in the conduction band,28 due to the low energy barrier for thermoionization of shallow traps. Therefore, if such shallow traps stabilized by Al3+

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

states, are the major type of the defects in TiO2, the lifetime of the charge carriers becomes longer and thus, the photoactivity of TiO2 samples becomes higher.

Figure 7. Scheme of the Fermi level shift in dependence on Al-concentration into TiO2. However, with increase of the Al-dopant concentration > 0.5 wt.%, the positive charge of oxygen vacancies can be compensated by dopant cations only forming the defect clusters such as / / × [ Al Ti − VOoo − Al Ti ] .

Clearly, the higher the dopant concentration, the more fully compensated

defect clusters are formed. At the same time, such clusters, Al stabilized oxygen vacancies, behave as deep traps. In this case, the Fermi level should be shifted towards the valence band as it is observed for all Al-TiO2 samples with Al content higher than 0.5 wt.%. Al stabilized oxygen vacancies acting as deep traps can be very effective recombination centers. Therefore, the stabilization of such defect clusters results in effective decay of the photogenerated charge carriers and consequently, in a significant decrease of the photoelectrochemical activity of Al doped TiO2. The transformation of the compensating defects from shallow traps as ineffective recombination centers to deep traps as effective recombination centers with increase of Al dopant concentration is responsible for the observed experimental dependencies with extrema of the work function and photoelectrochemical activity. Note, that similar behavior for Al doped TiO2 was previously observed for photocatalytic activity where optimal performance was also achieved for the photocatalysts with Al dopant concentration 0.5 wt.%.25 The existence of the optimal dopant concentration was predicted theoretically by Bloh29. However, the results of the present studies demonstrate that the model considered there should also take into consideration the possibility of the alteration of the type of compensating defects with the increase of the dopant concentration.

ACS Paragon Plus Environment

Page 14 of 19

Page 15 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Conclusions The results of the experimental studies of photoelectrochemical activity of Al-doped TiO2 as a function of the dopant concentration within the concentration range 0 – 1.1 wt.%, demonstrate the existence of the optimal Al concentration 0.5 wt.% for achieving maximal activity of TiO2 photo-anodes. The dependence of the photoelectrochemical activity on the Al dopant concentration correlates with the corresponding dependencies of the flat band potential and work function indicating a shift of the Fermi level toward the conduction band in the Al concentration range 0 – 0.5 wt.% and toward the valence band in the Al concentration range 0.5 – 1.1 wt.%. Such alteration of the thermodynamic characteristic can be explained in terms of alteration of the oo / type of major compensating intrinsic defects from [ VO - Ti Ti ] for Al concentration < 0.5 wt.% / oo / × acting as shallow traps, to [Al Ti − VO − Al Ti ] for Al concentration > 0.5 wt% acting as deep

traps. The transformation of the major type of compensating defects leads to an increase of the efficiency of charge carrier recombination. Thus, the increase of the Al dopant concentration affects both the number and the type of compensating defects, which results in the optimal dopant concentration for the higher photoelectrochemical activity of Al doped TiO2.

Acknowledgments. The present study was performed within the Project “Establishment of the Laboratory “Photoactive Nanocomposite Materials” No. 14.Z50.31.0016 supported by a Megagrant of the Government of the Russian Federation. We are also grateful to the RC “Nanophotonics”, RC “Nanotechnology”, RC “Chemical Analysis and Materials Research Centre”, RC “X-ray Diffraction Studies”, RC “Optical and Laser Materials Research” of the Research Park at the Saint-Petersburg State University for helpful assistance in conducting the synthesis and the characterization of the samples.

Supporting Information XRD patterns, XPS spectra of x-Al-TiO2 (x - 0.0, 0.1, 0.3, 0.5, 0.7, 0.9, and 1.1 wt.%), SEM images of 0.0-Al-TiO2 (a) and 0.5-Al-TiO2 electrodes, the dependences of photocurrent density on light intensity, current-voltage dependencies for x-Al-TiO2 (x - 0.0, 0.1, 0.3, 0.5, 0.7, 0.9, and 1.1 wt.%) electrodes under the dark and irradiation conditions, band gap energy estimation from

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

spectral dependencies of the photocurrent, Mott–Schottky plots of Al(x)-doped TiO2 electrodes, and the dependence of the flat band potential determined from the Mott-Shottky plots on the Aldopant concentration.

References: 1.

Fujishima, A.; Honda, K. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37−38.

2.

Nakata, K.; Ochiai, T.; Murakami, T.; Fujishima, A. Photoenergy conversion with TiO2 photocatalysis: New Materials and Recent Applications. Electrochim. Acta 2012, 84, 103−111.

3.

Fujishima, A.; Zhang, X.; Tryk, D. A. TiO2 Photocatalysis and Related Surface Phenomena. Surf. Sci. Rep. 2008, 63, 516−582.

4.

Henderson, M. A. A Surface Science Perspective on TiO2 Photocatalysis. Surf. Sci. Rep. 2011, 66, 185–297.

5.

Robertson, P.; Robertson, J.; Bahnemann, D. Removal of Microorganisms and Their Chemical Metabolites from Water Using Semiconductor Photocatalysis. J. Hazard. Mater. 2012, 211–212, 161–171.

6.

Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Visible-Light Photocatalysis in Nitrogen-Doped Titanium Oxides. Sci. 2001, 293, 269-271.

7.

Pelaez, M.; Nolan, N. T.; Pillai, S. C.; Seery, M. K.; Falaras, P.; Kontos, A. G.; Dunlop, P. S. M.; Hamilton, J. W. J.; Byrne, J. A.; O’Sheaf, K.; Entezari, M. H.; Dionysiou, D. D. A review on the visible light active titanium dioxide photocatalysts for environmental applications. Appl. Cat. B: Env. 2012, 125, 331–349.

8.

Nie, X.; Zhuo, S.; Maeng, G.; Sohlberg, K. Doping of TiO2 Polymorphs for Altered Optical and Photocatalytic Properties. Int. J. Photoenergy 2009, 2009, 294042.

9.

Choi, J.; Park, H.; Hoffmann, M. R. Effects of Single Metal-Ion Doping on the Visible-Light Photoreactivity of TiO2. J. Phys. Chem. C. 2010, 114, 783–789.

10. Livraghi, S.; Paganini, M. C.; Giamello, E.; Selloni, A.; Di Valentin, C.; Pacchioni, G.-F. Origin of Photoactivity of Nitrogen-Doped Titanium Dioxide under Visible Light. J. Am. Chem. Soc. 2006, 128, 15666−15671.

ACS Paragon Plus Environment

Page 16 of 19

Page 17 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

11. Kuznetsov, V. N.; Serpone, N. On the Origin of the Spectral Bands in the Visible Absorption Spectra of Visible-Light-Active TiO2 Specimens Analysis and Assignments. J. Phys. Chem. C 2009, 113, 15110−15123. 12. Ikeda, S.; Sugiyama, N.; Pal, B.; Marc, C.; Palmisano, L.; Noguchi, H.; Uosaki, K.; Ohtani, B. Photocatalytic Activity of Transition-Metal-Loaded Titanium (IV) Oxide Powders Suspended in Aqueous Solutions: Correlation with Electron Hole Recombination Kinetics. Phys. Chem. Chem. Phys. 2001, 3, 267−273. 13. Kuznetsov, V. N.; Emeline, A. V.; Rudakova, A. V.; Aleksandrov, M. S.; Glazkova, N. I.; Lovtcius, V. A.; Kataeva, G. V.; Mikhaylov, R. V.; Ryabchuk, V. K.; Serpone, N. Visible−NIR Light Absorption of Titania Thermochemically Fabricated from Titanium and Its Alloys; UV- and Visible-Light-Induced Photochromism of Yellow Titania. J. Phys. Chem. C. 2013, 117, 25852−25864. 14. Escudero, A.; Delevoye, L.; Langenhorst, F. Aluminum Incorporation in TiO2 Rutile at High Pressure: An XRD and High-Resolution

27

Al NMR Study. J. Phys. Chem. C. 2011, 115,

12196–12201. 15. Gesenhues, U. Al-doped TiO2 Pigments: Influence of Doping on the Photocatalytic Degradation of Alkyd Resins. J. Photochem. Photobiol. A: Chem. 2001, 139, 243–251. 16. Kuznetsov, V. N.; Ryabchuk, V. K.; Emeline, A. V.; Mikhaylov, R. V.; Rudakova, A. V.; Serpone, N. Thermo- and Photo-Stimulated Effects on the Optical Properties of Rutile Titania Ceramic Layers Formed on Titanium Substrates. Chem. Mater. 2013, 25, 170−177. 17. Lӧbl, P.; Huppertz, M.; Mergel, D. Nucleation and Growth in TiO2 Films Prepared by Sputtering and Evaporation. Thin Solid Films 1994, 251, 72−79. 18. Wang, H.; He, J.; Boschloo, G.; Lindström, H.; Hagfeldt, A.; Lindquist, S.-E. Electrochemical Investigation of Traps in a Nanostructured TiO2 Film. J. Phys. Chem. B. 2001, 105, 2529−2533. 19. Yang, S.; Guo, S.; Xu, D.; Xue, H.; Kou, H.; Wang, J., Zhu, G. Improved Efficiency of DyeSensitized Solar Cells Applied with F-Doped TiO2 Electrodes. J. Fluorine Chem. 2013, 150, 78–84. 20. Gionco, C.; Livraghi, S.; Maurelli, S.; Giamello, E.; Tosoni, S.; Di Valentin, C.; Pacchioni, G. Al- and Ga-Doped TiO2, ZrO2, and HfO2: The Nature of O2p Trapped Holes from a

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Combined Electron Paramagnetic Resonance (EPR) and Density Functional Theory (DFT) Study. Chem. Mater. 2015, 27, 39363945. 21. Ibadurrohman, M.; Hellgardt, K. Photoelectrochemical Performance of Graphene-Modified TiO2 Photoanodes in the Presence of Glycerol as a Hole Scavenger. Int. J. Hydrogen Energy 2014, 39, 18204−18215. 22. Liu, M.; De Leon Snapp, N.; Park, H. Water Photolysis with a Cross-Linked Titanium Dioxide Nanowire Anode. Chem. Sci. 2011, 2, 80–87. 23. Radecka, M.; Rekas, M.; Trenczek-Zajac, A.; Zakrzewska, K. Importance of the Band Gap Energy and Flat Band Potential for Application of Modified TiO2 Photoanodes in water Photolysis. J. Power Sources 2008, 181, 46–55. 24. Chen, Z.; Jaramillo, T. F.; Deutsch, T. G.; Kleiman-Shwarsctein, A.; Forman, A. J.; Gaillard, N.; Garland, R.; Takanabe, K.; Heske, C.; Sunkara, M.; McFarland, E. W.; Domen, K.; Miller, E. L.; Turner, J. A.; Dinh, H. N. Accelerating Materials Development for Photoelectrochemical Hydrogen Production: Standards for Methods, Definitions, and Reporting Protocols. J. Mater. Res. 2010, 25, 3–16. 25. Murashkina, A. A.; Murzin, P. D.; Rudakova, A. V., Ryabchuk, V. K.; Emeline, A. V.; Bahnemann, D. W. Influence of the Dopant Concentration on the Photocatalytic Activity: Al-Doped TiO2. J. Phys. Chem. C 2015, 119, 24695–24703. 26. Hu, T.; Wang, Z.; Ma, N.; Du, P. Control of VO•• ∼ TiTi′ Dipole Pairs as well as MgTi″ Defects on Dielectric Properties of Mg Doped (Pb0.35Sr0.65)TiO3 Thin Film. J. Appl. Phys. 2016, 119, 014103. 27. Di Valentin, C.; Pacchioni, G.; Selloni, A. Reduced and n-Type Doped TiO2: Nature of Ti3+ Species. J. Phys. Chem. C. 2009, 113, 20543–20552. 28. Gesenhues, U.; Rentschler, T. Crystal Growth and Defect Structure of Al3+-Doped Rutile. J. Solid State Chem. 1999, 143, 210–218. 29. Bloh, J. Z.; Dillert, R.; Bahnemann, D. W. Designing Optimal Metal-Doped Photocatalysts: Correlation between Photocatalytic Activity, Doping Ratio, and Particle Size. J. Phys. Chem. C. 2012, 116, 25558−25562.

ACS Paragon Plus Environment

Page 18 of 19

Page 19 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

TOC Graphic

ACS Paragon Plus Environment